The Lost Xbox 1.5 Plan

[ban25]

If there was a postmortem article, sticking with x86/PC architecture would be listed as a #1 decision not to repeat. A multicore PowerPC-based successor was the focus of pretty much everyone from 2002 onward and it certainly went without saying that no one was going to divert efforts to an intermediate console, especially one that would be cursed with the "decision not to repeat" attribute.

Was this strictly a business/IP licensing decision? I think most people expected Microsoft to continue with x86 to maintain backwards compatibility and similarity with the PC platform. Perhaps when things did not go well with the NVIDIA license deal and when they began looking at other graphics vendors, they decided to look further afield for a CPU partner as well. Or visa versa.

 

[DudeWheresMyXbox]

Business, IP and better engineering reasons all lined up.

Significant cost and size reduction in later years can be accomplished through ongoing silicon optimization. Just look at PS2 EE+GS. Off-the-shelf parts, especially from hardware companies that quickly move on to their next product, often don't include that benefit.

Ability to add important silicon features game developers will benefit from that can be supported at a near-automatic level by the compiler and OS.

Lots of performance reasons.

The security of conventional x86 architecture is fundamentally exploitable.

Contract arguments with NVidia made good corporation soap opera for observers, but spite doesn't lead to good technical or business decisions. Many options were evaluated for Xbox 360's hardware outside of CPU family.

I agree that a lot of people have been surprised when the strategies MS made up in a hurry to get a product out to market are changed for a better-planned successor.

 

[Fox5]

I call shens on Xbox 1.5. Considering that backwards compatibility on 360 generally doesn't improve graphics, I doubt a 1.5 with better graphics was in the works.

Oh, and x86 was probably a bad idea for a few reasons....
1. AMD and Intel don't license their designs.
2. AMD and Intel are already fab constrained with the PC market.
3. The aim of their cpu designs may not match what Microsoft wanted. Low flops and AMD's integrated memory controller is far too PC focused to make good sense in a console.
4. Probably more that can fairly easily be pointed out.

 

[Megadrive1988]

I'd prefer to keep my job, thanks.

And would anyone really believe that your average PR type knows about IGN, let alone B3D? I sure don't


Contract signing took a long, long, long time.

If there was a postmortem article, sticking with x86/PC architecture would be listed as a #1 decision not to repeat. A multicore PowerPC-based successor was the focus of pretty much everyone from 2002 onward and it certainly went without saying that no one was going to divert efforts to an intermediate console, especially one that would be cursed with the "decision not to repeat" attribute.


I'll admit that my understanding of your post was incorrect. The "E3 mockup" part threw me off.

with the IBM-Microsoft deal being announced in 2003, it was probably a done deal in late 2001 if not 2002. who knows, maybe it was done before the first Xbox was completed, but I have no doubt in my mind it was done long before 2003.


heck, the ink on the Xbox3 (360's successor) contract might already be drying now.

 

[Megadrive1988]

I'm not sure what you're disagreeing with as it sounds like you agreed to me.

awww don't worry about it -- i basicly skimmed over your post -- plus i had a hangover still

 

[3dcgi]

with the IBM-Microsoft deal being announced in 2003, it was probably a done deal in late 2001 if not 2002. who knows, maybe it was done before the first Xbox was completed, but I have no doubt in my mind it was done long before 2003.


heck, the ink on the Xbox3 (360's successor) contract might already be drying now.

Don't underestimate the ability of business people to squabble over money. Even the richest company in the world takes a long time to agree on a contract. A design can be created before the contract is signed, but full blown implementation is sure to be delayed until there is a guarantee.